Acknowledgements. This work would not have been possible without the help of many individuals and institutions whose support is now most gratefully acknowledged. Our thanks are due in the first place to several members of the Greek Archaeological Service for help before and during the field-work and subsequent museum studies: Mr. D. Lazaridis, Ephor of Attica in 1966; Mr. A. Liangouras, then Epimelete; Mrs. Helene Theodoraki, the Service representative on the site; Mr. E. Mastrokosta, Ephor in 1969; and the staff of the Peiraeus Museum. Our thanks go also to Mr. Chrysostomos I. Neamonitakis of the Organismos Dioikeseos Ecclesiastikes Periousias, representing the owners of the site. We are much indebted to former Directors of the British School at Athens, Mr. A. H. S. Megaw whose indefatigable labours in 1966 in connection with the preliminary arrangements made the project possible, and Mr. P. M. Fraser who facilitated arrangements for the final tests on materials in 1969. For encouragement and generous financial aid we owe a debt of gratitude to the University of Cambridge Faculty of Classics, and in particular to Professor R. M. Cook and Mr. A. G. Woodhead, the then Secretary; to the Oxford University Craven Fund; to the University of Manchester and in particular to the then Vice-Chancellor, Sir William Mansfield Cooper; and to the University College of North Wales and in particular Professor R. E. Wycherley. The help of several members of the American School of Classical Studies and Agora Excavation staffis acknowledged: Professor Homer Thompson very kindly visited our site to discuss the remains and finds with us; Professor E. Vanderpool gave encouragement and advice; and Mr. T. Leslie Shear, Jr., Miss Virginia Grace, the late Miss Lucy Talcott, and other members of the Agora Staff gave us kind reception, advice on pottery studies, and permission to examine material in the Stoa of Attalus and to quote parallels.

Our debt to fellow members of the British School is also great: to Mr. K. MacFadzean of Glasgow University for checking our site survey and taking levels; to Dr. M. P. Price of the British Museum for help in identifying the coins; to Dr. B. A. Sparkes of Southampton University for freely giving much expert advice on the pottery material and (with the late Miss Lucy Talcott) generally allowing us to consult Agora xii in proof; to Mrs. R. Ellis Jones for undertaking the catering for the whole excavation team at Vari; and Miss P. Rhodes for acting as relief driver. Local arrangements were helped also by Mr. G. Tsirikos and Mr. J. Katsikis of Vari.

In obtaining in 1969 samples from our ‘combed’ pottery for scientific analysis we were greatly helped by Mr. P. Siphnaios, Mr. Chalkias of the British Council Centre at Athens, Mr. A. Stavropoulos of A. E. Bioryl of Kephissia, and by the following members of the Athens Polytechnic University: Professor N. I. Theophanopoulou, of the Engineering Department, Professor E. Kambouris of the Chemical Technology Department, and in particular by Dr. K. Th. Bellos of the latter department who conducted the preparation of the solvent samples at the Peiraeus Museum and in his laboratory. For particular help with these tests and with the pottery material discussed in the Excursus on the Combed Ware, we acknowledge various debts of gratitude to Dr. J. D. Bu'Lock, Reader in Microbial Chemistry at the University of Manchester; Dr. Eva Crane, Director of the Bee Research Association; Mr. Harold Inglesent, past President of the Manchester Beekeeping Association; and Professor F. R. Jevons, Professor of Liberal Studies in Science at Manchester University. For permission to produce photographs of beehives and beekeeping practices we are obliged to Professor O. Broneer, Director of the American Excavations at Corinth (for the Orestada vessel); the Bee Research Association Picture Library and Brother Adam, Buckfast Abbey (for the old straw hive, skep, and eke, and the modern Cretan beehives); and Dr. Nora E. Scott, Curator of the Department of Egyptian Art, Metropolitan Museum of Art, New York (the scene from the Rekhmire tomb); and the University of Chicago Press (the scene from the tomb of Pabes). In this connection we place on record our gratitude to Mr. M. I. Geroulanos of Trachones, Attica, for his gracious reception of us on several occasions, for showing us his collection of antiquities and discussing his ‘combed’ material, providing photographs of the same, and writing a parallel article on the Trachones examples for this volume (here Appendix I), to supplement our Excursus.

¹ For the earlier work on the Dema House see BSA lvii (1962) 75–114; Archaeology xvi (1963) 276–83. Brief notices of the work at Vari, with plan and photographs, appeared in the British School's and Hellenic Society's Archaeological Reports 1966—7, Archaeology in Greece 5, also in Antiquity xli (1967) 126 pl. 1 oc; A Delt xxii (1967) B Chronika 136–7 pl. 104; AJA lxxi (1967) 296; BCH xci (1967) 626–7. A general account of the field-work and studies, with a summary of the results, is given in Jones, J. E., The Greeks (1971) 33–52Google Scholar (The Young Archaeologist Series, Rupert Hart-Davis Educational Publ., London and G. P. Putnam's Sons, New, York).

² We could find no name given to the ridge site itself. The local names are not helpful. According to Dodwell the mountain where the Cave of Pan is located was called Rapsana in the early nineteenth century, and that name appears again on the map in Leake's Athens and the Demes of Attica ii, but not as the peak nearest to the site. Curtius and Kaupert, Karten von Attika VIII, give Krevati as the name of a major peak north-east of the site. In the early years of this century, according to the American excavators of the Cave of Pan, the location was called Kapsala, or was often named then, as now, Spelaion or Spelaion Nymphon after the cave itself. According to a local informant the names Arseni and Dousas are given to the lower and upper ends of the enclosed valley on the west side of the house site.

³ The site could be approached from Vari on foot by a dirt track leading directly north-east along the northern limit of the plain to the point where a narrow valley breaks the line of hills, or by car from the Vari-Koropi road, by turning off just before the second bridge on to a track following a stream-bed to the same point (PLATE 63); from there the track continued up an enclosed valley, passing the ridge site of the house. But at the time of going to press (1972) both approaches have been blocked, through the enclosure of private plots of land east of Vari village and in particular by the fencing-in of a large tract of plain and hillside immediately south of the site to form a military practice area; new roads are being constructed within this.

⁴ In 1805 Dodwell visited the Cave of Pan ascending ‘through a forest of small pines’, and again mentions descending through a ‘pine forest’ (and observing at the foot of the hill ‘several vestiges of antiquity’, which may or may not have been the house site). Air photographs taken in 1945 show the hills largely denuded of trees, so the present cover is of fairly recent growth.

⁵ The cave was visited in 1765 and first described in print by R. Chandler in his Travels in Greece i 169 ff.; for their early descriptions see Dodwell, E., A Tour through Greece i (1819) 550–5Google Scholar; the Revd. Wordsworth, Chr., Athens and Attica (1836) 192–201Google Scholar; and Leake, W. M., Athens and the Demes of Attica ii (1841) 57 ff.Google Scholar The cave was excavated by members of the American School of Classical Studies at the beginning of this century; AJA vii (1903) 263–349.

⁶ Coastal Demes 38 fig. 3 (plan of Vari area) and 44 (description); the site is one of three ‘farm-houses’ sited 2 km. or more outside Vari, between the Koropi road and the mountains. Eliot discusses them in terms of a modern exochi, a temporary home among the fields used seasonally by farmers normally residing in a central village.

⁷ This hut seemed to have been a huntsman's cabin; traces of an outside hearth and camping debris showed that it was used for occasional brief stays, rather in the manner suggested by Eliot for the ancient site itself. The hut was completely demolished in 1967, but a replacement was built by 1968 further uphill, under the west side of the terraced foundation of the ancient house; this incorporated loose stones cleared from the house and dumped on the slope below. Both exemplify the kind of brief reoccupation of ancient sites which in the case of an early re-use would have left some finds that might be regarded as contamination or secondary archaeological material.

⁸ See pp. 363 and 364 below, and pp. 415 ff. for a discussion of the date.

⁹ See pp. 360 above and 415 f. below. The survival of these sherds does not totally preclude a re-use of this area at some later date, for they lay under or close to the excavators' barrow-run and yet, filled and surrounded with earth, survived without further damage. But they suggest that any reoccupation extending to this area would have been at a level well above their own position, and so high above the floor and ‘fittings’ of the room that the room itself could not have been a re-usable unit.

¹⁰ Pottery catalogue 31, 60, 93, 128.

¹¹ A similar pot-hole was noted in the rocky surface 2 m. outside the north wall of the house and was cleared out. It was deeper but contained only dark humus and small stones. The shallower hole in room II probably represents the lower end of a pot-hole disturbed and reduced in effective depth by the levelling of the rock within the room, and then replugged with some of the earth used to lay the floor.

¹² See pp. 360 above and 415 f. below; catalogue 4, 5, 13; also a few coarse sherds.

¹³ See p. 372 below; coin 1.

¹⁴ See p. 373 below; coins 3 and 4.

¹⁵ The end of the Vari House might well have been very like that of a number of isolated farmsteads and houses and huts in the Peneios Valley, Elis. Because land there was to be flooded to create an irrigation reservoir many such properties had to be given up, and on successive visits in 1969 were observed the stages of abandonment and partial dismantling of houses and huts and the general salvage of movable or re-usable material. Timber doors, windows, frames, roof-beams, and other fittings were removed; roof-tiles were stacked to be carted away and mud-brick walls were just left standing or half thrown down for the weather to complete their dissolution. Afterwards, there would be left only a few beams too rotten for re-use, a few fragments of tiles broken when removed from the roof or carelessly stacked, unwanted rubbish, much of it perishable, and the walls, stone foundations, and perhaps a cement floor; such would be the raw material of the ‘archaeological evidence’ for what had been seen a few months earlier as a standing and occupied house. There is some ancient literary evidence for the removal of building-material, legal and illegal. Lysias xix (On the property of Aristophanes) 31 notes the rapid disappearance of valuable materials from houses standing empty after confiscation by the state: ἀλλὰ τόδε σκοπεῖτε τῶν ἄλλων, ὅσων ἐδημεύσατε 〈τὰ〉 χρήματα, οὐχ ὅπως σκεύη ἀπέδοσθε, ἀλλὰ καὶ αἱ θύραι ἀπὸ τῶν οἰκημάτων ἀφηρπάσθησαν ἡμεῖς δὲ ἤδη δεδημευμένων καὶ ἐξεληλυθυίας τῆς ἐμῆς ἀδελφῆς φύλακα κατεστήσαμεν 〈ἐν〉 τῇ ἐρήμῃ οἰκίᾳ, ἵνα μήτε θυρώματα μήτε ἀγγεἰα μήτε ἄλλο μηδὲν ἀπόλοιτο. Again, Thucydide (ii. 14. i, ii. 17. 1 and 3) mentions that many Athenians, when evacuating their families and stock from their farms and removing to Athens in 431 B.C., even pulled down the woodwork of their houses and brought the material along with them, partly perhaps to protect it and partly to re-use it for building shacks in the ‘refugee shanty town’ that developed in the open spaces of Athens and the corridor of the Long Walls. Ancient tenancy agreements sometimes stipulated that a tenant was to provide his own doors, woodwork, and tiles, etc., and could remove them at the end of his term, provided he had satisfied the property-owners, but would forfeit these elements if he were unsatisfactory: IG ii². 2499. 11–14, 30–7 (306/5 B.C.); Olynthus xii. 449–50; cf. Hesperia xvii (1948) 293.

¹⁶ BSA lvii (1962) 88, 100.

¹⁷ For the attribution of the name ‘lopas’ to the pots long called ‘casseroles’ see Amyx, , Hesperia xxvii (1958) 210 n. 76Google Scholar; Sparkes, , JHS lxxxii (1962) 130.Google Scholar

¹⁸ The choice of name is difficult. Pseudo-analogies from the modern world seem inappropriate (as ‘umbrella-stands’; see, e.g., McCredie, Hesperia Supp. xi. 46 no. 5). The vessels are not very like the kalathos in fine ware (see Richter, G. M. and Milne, M. J., Shapes and Names 13 f. pl. 90Google Scholar). The shape is, however, generally similar to the ancient wool-basket called a kalathos (see ibid.), so we are content to follow Corbett, , Hesperia xviii (1949) 335 no. 95.Google Scholar

¹⁹ Of our examples only one, 150, has a ring foot. Agora xii no. 1853 (n. 21 below, no. 1) might also seem to be exceptional, if we follow Corbett. But, although the two authorities give different descriptions, the base is in fact so low that its fundamental character is that of a flat base, as is apparent from the photographs. It should be noted, however, that though the bases are theoretically flat they are not perfectly so, but tend to have a slightly convex shape, and their junctions with the walls of the pot are curved.

²⁰ The height is not known for certain in any of our examples, but see below.

²¹ 21 We collect in this footnote all the parallels to our coarse kalathoi that we know of. They are presented in chronological order, as far as possible, first the published examples, and then the unpublished. Brief information about provenance, characteristics, and date is given, so far as we have been able to obtain it.

²² See n. 19 above.

²³ See Agora xii. 217 f.

²⁴ See Broneer, , Hesperia xxvii (1958) 32 no. 42, and pl. 14bGoogle Scholar; Kardara, , AJA lxv (1961) 264 f.Google Scholar, and pl. 81 fig. 6. The fabric is Corinthian. Several fragments of similar vessels were found on the same site (Kardara, ibid.).

²⁵ We are grateful to Mr. Charles Williams, who kindly examined the pot on our behalf, for these details.

²⁶ Kardara, op. cit. 263.

²⁷ See Anderson, , BSA xlix (1954) 137 no. 28Google Scholar; 142 no. 79; 173 fig. 5. No. 79 has an est. D. at mouth of c. 0·40 m.

²⁸ See Varoucha-Christodoulopoulou, , AE 1953/1954, Part iii (Athens, 1961), 335 no. 11; fig. 6Google Scholar, ε.

²⁹ Ibid.

³⁰ See n. 21 above, no. 8.

³¹ See pp. 446; PLATES 83–5.

³² We list them here (as in n. 21 above) in chronological order within two groups, published and unpublished, adding in the case of the latter such information as we have acquired about characteristics, provenance, and date.

³³ See n. 21 above, nos. 4–6.

³⁴ See n. 24 above.

³⁵ There is a short, well-documented discussion in Agoraxii. 217 f.

³⁶ See n. 21 above, no. 11.

³⁷ AJA lxv (1961) 264f.; see also above n. 24.

³⁸ AE 1953/4, part iii (Athens, 1961) 335.

³⁹ Aeora xii. 217 f.

⁴⁰ See Crane, E., ‘The World's Beekeeping—Past and Present’, ch. I (pp. 1–18) of The Hive and the Honey Bee, ed. Grant, R. A. (1963).Google Scholar In an older work, Feburier, , Traité complet théorique et pratique sur les abeilles (Paris, 1810) 142–4Google Scholar, the account of terracotta beehives, which describes in particular those in use in Greece, is most suggestive when compared with our material.

It is appropriate at this point to acknowledge our very great debt to Dr. Eva Crane, Director of of Bee Research Association, who has not only kindly discussed our pottery with us on several occasions, but has also generously given us access to the rich material on beekeeping and its history in the library of the Bee Research Association. Since Dr. Crane also took the trouble to read a draft of this Excursus and made a large number of most valuable observations, it is all the more necessary to emphasize that she bears no responsibility for any statement in it unless specifically attributed.

Mr. Harold Inglesent, past President of the Manchester Beekeeping Association, has also been most helpful in many discussions, generously giving us the benefit of his great knowledge and experience of beekeeping.

⁴¹ See Crane, ibid., especially p. 3 fig. 3.

⁴² See, e.g., Strabo ix. 399; Pliny, , NH xi. 13. 32Google Scholar; Paus. i. 32. i.

⁴³ Plutarch, , Solon xxiii. 6.Google Scholar

⁴⁴ See above, n. 21, nos. 4–8. However, no. 4, from the Kynosoura peninsula, probably emanated from a settlement rather than a military camp; see McCredie, Hesperia, Supp. xi. 45 f. with fig. 8 (on site F). Nor are the sherds of no. 8 necessarily from the military occupation of Helioupolis. The late kalathoi from the Isthmus, no. 11, were also found in a fortress.

⁴⁵ Cf. the remarks on Koroni, , Hesperia xxxi (1962) 38 n. 7.Google Scholar

⁴⁶ Works on ancient beekeeping: Malcolm Fraser, H., Beekeeping in Antiquity (London, 1931; 2nd ed. 1951)Google Scholar; Klek, J. and Armbruster, L., Die Bienenkunde des Altertums, published successively in Archiv für Bienenkunde i no. 6 (1919)Google Scholar, ii no. 7 (1920), and iii no. 8 (1921), (hereinafter Klek and Armbruster 1919, 1920, 1921). RE s.v. Bienen, Bienen zucht (also in Supp. iv); Daremberg et Saglio, Dictionnaire des Antiquités, s.v. Apes (a useful article, though it contains many of the false statements about ancient hives which have found their way into general books on the history of beekeeping); Forbes, R. J., Studies in Ancient Technology v (2nd ed.Leiden, 1966) 80–111Google Scholar, is a rich collection of material. Owing to the generosity of the staff, we were also able to consult an M.A. thesis on Ancient Beekeeping prepared by Thomas Rigley in the Classics Department of the University of Leeds (1951).

⁴⁷ The supposed Roman relief which includes the wicker beehive illustrated in the Daremberg et Saglio article (see previous note) is given in de Montfaucon, B., L'Antiquité Expliquée (Paris, 1729) i pl. 204Google Scholar, but has long been known to be a forgery. See CIL vi 5. 3153^* and vi. i pp. lv f. (no. lix).

⁴⁸ See the discussion by Kueny, G. in Journ. Near East. Studies ix (1950) 84–93.CrossRefGoogle Scholar A fourth scene, of the Eighteenth Dynasty, is now ‘scarcely visible’; see Lucas, A., Ancient Egyptian Materials and Industries, 4th edition revised by Harris, J. A. (London, 1962) 26.Google Scholar We are much indebted to Dr. A. F. Shore for introducing us to the material from Egypt.

⁴⁹ The material is clearly indicated by the colour in the painting from the tomb of Rekhmirē: ‘clay (purplish gray)’; see Davies, Norman de Garis, The Tomb of Rekhmirē at Thebes (New York, 1944) 45.Google Scholar We are not convinced that we can distinguish two types of hives in these paintings, as Kueny 88 f., though those of Rekhmirē are more cylindrical and less elongated than those of the tomb of Pabes.

⁵⁰ The hives in the painting from the tomb of Pabes are represented very schematically, and those from the tomb of Neuserrē are damaged, but the arrangement of those from the tomb of Rekhmirē is very clear, and the interpretation is quite definitely established by the presence of a bee and the ‘smoker’; see Davies pl. 49. Thus there is no question that in all three an arrangement of hives in a pile or heap is depicted.

⁵¹ See above, p. 402.

⁵² Illustrated in MA xii (1902) 54, fig. 15, and xiv (1904–5) 447, fig. 62. See now Pernier, L., Banti, L., Palazzo minoico S Festos ii. 81 f.Google Scholar figs. 38, 39; 91 fig. 45; 93 fig. 46. The original interpretation as a beehive (see e.g. Mosso, Angelo, Palaces of Crete (London, 1907) 154Google Scholar) was repeated in books on the history of beekeeping; e.g. Ransome, Hilda, Sacred Bee (London, 1937) 63.Google Scholar

⁵³ For this beehive see Donaldson, T. L., Pompeii (London, 1827). ii.Google Scholar and plate opposite p. 12, from which all the later works which refer to it (e.g. Clarke, W., Pompeii (London, 1846) ii. 79Google Scholar; Billiard, R., L'Abeille et L'apiculture (Lille, 1900) 43Google Scholar; Daremberg et Saglio, art. cit.) seem to derive. Donaldson gives no information as to the circumstances in which it was found, nor does he discuss the identification, merely stating that it is of bronze, was found at Pompeii, and is in the Museum at Naples. Dr. Crane has ascertained that the museum authorities at Naples do not now know of its existence, and its absence from such catalogues as Museo Borbonico (1824–67) and that of Roux, H. of 1861 (Herculaneum et Pompeii, Recueil general etc., Paris, 1861–1877)Google Scholar, not to mention later works of the same kind, suggests that it disappeared long ago. We are also indebted to Dr. Crane for the expert opinion that it cannot have been a beehive: in addition to the unsuitable and unparalleled material, the rows of holes and horizontal compartments could not serve such a purpose.

⁵⁴ See Mr. and MrsCunnington, B. H., Wiltshire Archaeological Magazine xxxviii (1913) 53–105Google Scholar, especially p. 100, and pls. 4 nos. 1 and 4, pl. 6 no. 5. We are grateful to Mr. A. M. Burchard of Devizes Museum for helpful information about this pottery.

⁵⁵ For the dating of similar pottery and sites see Brailsford, J., Proc. Prehistoric Soc. xxiv (1958) 101–19.CrossRefGoogle Scholar

⁵⁶ The identification as a beehive was proposed by the excavators in their original publication (see n. 54), and was also made independently by Dr. Crane when she visited Devizes Museum.

⁵⁷ See the works cited above in n. 46, especially those of Fraser and of Klek and Armbruster, both of which constitute complete surveys of the ancient literature on beekeeping. The introduction to Klek and Armbruster (1919. 3–10) is a useful brief survey of the ancient literature, both lost and extant.

⁵⁸ By Aristomachus of Soli (who devoted fifty-eight years entirely to beekeeping, and wrote a work entitled Melissour-gika) and Philiscus of Thasos; see Pliny, , NH xi. 8. 19.Google Scholar Cf. RE s.v. Aristomachus no. 20; Klek and Armbruster, 1919. 7.

⁵⁹ Aristotle, HA v. 21–2, 553a–554b; ix. 40, 623b–627b; GA iii. 10, 759a–761a. Varro, , RR iii. 16.Google Scholar Columella ix. 2–16. 1. This is not to be taken as implying a low opinion of Virgil and Pliny the Elder, both of whom are well and sympathetically treated by Fraser (see n. 46 above). It is merely that they rarely provide information not already to be found in our earliest systematic source, Aristotle, and the practical accounts of Varro and Columella.

⁶⁰ For the character of Varro's and Columella's work cf. Klek and Armbruster, 1920. 2 ff.; 1921. 1 ff.

⁶¹ Varro, , RR iii. 16. 15Google Scholar; Columella, ix. 6. 1–2.

⁶² Varro, iii. 16. 17; Columella, ibid. Note that Fraser (46 f.) is in error when he says that Varro advises that terracotta hives be smeared with cow-dung to improve their thermal qualities. It is the plaited hives that Varro says should be treated thus, in order to make them smooth.

⁶³ Columella refers frequently to his main authorities, Virgil, Hyginus, and Celsus. Of these Hyginus (who wrote in the latter part of the first century B.C.; RE s.v. Iulius no. 278, 628 f.) is expressly praised for collecting the opinions of old authorities on beekeeping. See Columella, ix. 2. 1; cf. Klek and Armbruster, 1921. 4. Varro is very largely dependent on his literary forerunners, of whom a large number, Latin, Greek, and Carthaginian, can be listed; see H. Dahlmann, RE Supp. vi, s.v. M. Terentius Varro, 1199 ff.

⁶⁴ We have to reckon not only with the Greek influences, but also with Carthaginian, when we are dealing with sources for Roman writers on agriculture; see Dahlmann, ibid.

⁶⁵ See n. 59 above.

⁶⁶ See During, I., Aristoteles (Heidelberg, 1966) 506–8.Google Scholar However, there are dissenters; see the discussion by A. L. Peck in his Introduction to the Loeb edition of HA, liii ff.

⁶⁷ See During, loc. cit.

⁶⁸ Thus we cannot share the opinion of Klek and Armbruster (1919. 54) that the author of HA ix. 40 was a beekeeper. In both the truly Aristotelian passages and ch. 40 of HA ix the references to beekeepers (their knowledge and practices) make it quite clear that the author is not of their number; see e.g. GA iii. 760a. 1 ff.; HA ix. 626a. iff., 626b 1 ff., 627a. 15 ff.

⁶⁹ e.g. Platt, n. i to GA iii. io in his translation in the Oxford series, ‘… it is not to the credit of any modern writer that he should assert that the Greeks were well acquainted with the internal economy of the hive’.

⁷⁰ Compare the more cautious and more complicated discussions of the matter in Fraser 18 f., 106 f.

⁷¹ HA 624b. 8ff.

⁷² HA 623b. 26 ff.; 625a. 11 ff.

⁷³ On the ancient beekeepers’ desire to encourage swarms cf. Fraser 39.

⁷⁴ GA 760b. 8 ff.

⁷⁵ HA 624a. 30 ff.; cf. GA 759b. 9 ff., 36 ff.

⁷⁶ HA 625a. 16 ff.

⁷⁷ HA 554a. 15 ff.; 624b. 11 ff.; 625a. 6 ff.; 625b. 30 ff.

⁷⁸ As Fraser 18.

⁷⁹ Cf. Fraser 19.

⁸⁰ See especially HA 553a. 16 ff.

⁸¹ HA 623b. 18 ff.

⁸² See above, p. 403.

⁸³ διὸ καὶ πλέκουσί τινες περὶ τὰ σμήνη ὥστε τὰς μὲν μελίττας εἰσδύεσθαι, τοὺς δὲ κηφῆνας μὴ διὰ τὸ εἴναι αὐτοὺς μείӡους. HA 553b. 12 ff.

⁸⁴ Though Dr. Crane informs us that the old idea that numerous drones reduced the honey yield has, in fact, recently been shown to be mistaken.

⁸⁵ Cf. Fraser 21.

⁸⁶ ἀναπλάττουσι δὲ ὁτὲ μὲν καὶ αὐτὰ καθ᾿ αὑτὰ τά κηρία τὰ τω-ν κηφήνων, ὡς ἐπὶ τὸ πολὺ δ᾿ ἐν τοῖς τῶν μελιττῶν διὸ καὶ ἀποτέμνουσιν. HA 624b. 18 ff.

⁸⁷ Cominella ix. 11. 4; cf. Fraser 66 f.

⁸⁸ Ibid.; cf. Columella ix. 7. 2; 15. 11. Our PLATES 79, 80 show modern terracotta beehives in Cyprus, laid horizontally and open at both ends, with clay lids to plug the holes.

⁸⁹ Fraser, loc. cit.; cf. also 70 f.

⁹⁰ Pliny, , NH 16. 49Google Scholar; cf. Fraser 85.

⁹¹ See especially HA 553a. 29 ff., where Aristotle states that some called the rulers ‘mothers’, and the whole discussion in GA iii ch. 10.

⁹² See the statement, justly famous for the light it throws on Aristotle's methods, at GA 760b. 30 ff.

⁹³ In this conclusion we differ from so learned a beekeeper as Dr. Fraser; see for example his p. 106. However, he is also frequently led to remark that some statement in ancient literature on beekeeping implies primitive hives; see, e.g., pp. 19, 48, 70 f.; so that he is conscious of apparent contradictions; see, e.g., p. 67. One point is clear: if there were beehives with movable combs in Aristotle's day they were not like our kalathoi. The bar-hive of modern times was used in Greece well before the discovery of the true sectional hive (see Fraser 106 f.), but the kalathoi are the wrong shape to come into consideration as possible bar-hives. In shape the more bucket-like Orestada vessel (see above p. 399) looks more suitable, so long as its wide mouth was at the top (cf. the similar modern Cretan hive in PLATE 78d); but the combing of its interior could be held to show that it was expected that the combs would be fixed to the side of the vessel.

⁹⁴ Cf. the large numbers of theories to which Aristotle refers; e.g. HA v. 21.

⁹⁵ We owe the original suggestion that we should apply such a test to our material to Professor F. R. Jevons, Professor of Liberal Studies in Science, University of Manchester.

⁹⁶ It proved impossible to obtain a sample from the coarse lids, as the fragments were too large to insert into the glass flasks used for soaking. Similarly only a small selection of sherds from the terracotta rings could be inserted in the flask because of their size and curvature. Thus, in considering the negative result from sample (2), we need to bear in mind that this sample was taken from a smaller quantity of sherds than sample (1) from the coarse kalathoi. It is also necessary to point out that all the sherds we tested had been carefully cleaned, along with our other pottery, by soaking in dilute hydrochloric acid and thorough washing. It seems most desirable that the test we have carried out on our sherds should be repeated on other examples of pottery of the same type which have not been energetically cleaned beforehand.

⁹⁷ See above, pp. 400 f.

⁹⁸ Cf. above, n. 40.

⁹⁹ See above, p. 404 with n. 54.

¹⁰⁰ ix. 7. 5.

¹⁰¹ See above, p. 399.

¹⁰² Cf. above, n. 40.

¹⁰³ Cf. Kardara, , AJA lxv (1961) 265Google Scholar, who notes in her discussion of the Orestada vessel that one of the advantages of such pottery beehives is the possibility that they can be extended by a ‘bottomless cylindrical stem’ (although such extensions are not among the published material from the Rachi).

¹⁰⁴ See above, p. 402 and PLATE 79d.

¹⁰⁵ See above p. 403.

¹⁰⁶ Cominella ix. 7. 2; 15. 11; cf. Varro, , RR iii. 16. 16 f.Google Scholar

¹⁰⁷ Wg owe this knowledge to the expert advice of Dr. Crane.

¹⁰⁸ Cf. Columella ix. 15. 8, and Pliny, , NH xi. 23.Google Scholar Both mention various shapes of comb, which were determined by the different shapes of hives. Unfortunately neither author specifically relates the shape of the comb to a detailed description of a hive.

¹⁰⁹ See above, p. 403; Kueny, , Journ. Near East. Stud. ix (1950) 90 f.Google Scholar

¹¹⁰ The two analogous rings of later date (see above, p. 399) are considerably higher: 0·102 m. They might just have held three combs. It seems possible that clay or some such material might have been used to make a good joint between eke and hive and eke and lid; and if that were so there would have been in fact a little more space for comb-building than the exact height of the eke.

¹¹¹ xi. 15. 45.

¹¹² ix. 399.

¹¹³ See above, p. 404.

¹¹⁴ Agora xii. 217 f.

¹¹⁵ Cf. Fraser 95 f.

¹¹⁶ See the references in n. 42 above.

¹¹⁷ See FIG. 1 and, for a description of the cave, AJA vii (1903) 263ff.

¹¹⁸ Aelian, , VH x. 21Google Scholar; cf. Pliny, , NH xi. 18. 55Google Scholar; Cic, De Divinatione i. 36. 78; cf. Val. Max. i. 6. Ext. 3.

¹¹⁹ ‘The two huts built in recent years illustrate the point well enough. The earlier one took advantage of the ridge site, being set clear of the arable lower slopes, but was just outside the south-east corner of the outer enclosure (p. 359 above). The later one has been built outside the ancient house, but backs against its high west terrace foundation and re-utilizes some of its building-material, namely stones from the dump made on the side of the site.

¹²⁰ AJA vii (1903) 284 f. Cf. also Agora vii. 63: ‘It appears that A.D. 350–360 was the date of the resumption of regular visits to the cave.’

¹²¹ But the Byzantine coin from the fill of the south-west corner room could suggest that the still prominent walls of that room may have been re-used as a base for a hut or other shelter.

¹²² Contrast, for example, the House of Mikion and Menon in the Agora, Athenian, Hesperia xxxviii (1969) 383Google Scholar: ‘It soon became clear in the course of excavation that the building had enjoyed occupation for about a century and three-quarters. Undisturbed stratification of no less than five floors could be recovered from several parts of the house. On the basis of pottery found in these well-stratified layers, the chronology of the building can be established from its construction in the second quarter of the 5th century B.c. to its destruction at the end of the 4th century. During this period the building underwent various structural modification on at least three different occasions.’

¹²³ Hesperia xxxi (1962) 26–61.

¹²⁴ See Excursus.

¹²⁵ See Edwards, , Hesperia xxxii (1963) 109–11CrossRefGoogle Scholar; Grace, ibid. 319–34; the excavators’ reply, Hesperia xxxiii (1964) 69–75; and, more recently, McCredie, , Hesperia, Supp. xi (1966) 14 f.Google Scholar

¹²⁶ Hesperia iii (1934) 311–480.

¹²⁷ See for example no. 1632 p. 351 (= Thompson A 37).

¹²⁸ See McCredie, loc cit.

¹²⁹ Even if, as Miss Grace emphasized, op. cit., the attribution of Ptolemaic coins is very difficult; cf. Head, HN ² 846 and Amer. Num. Soc. Museum Notes vi (1954) 69.

¹³⁰ Cf. Grace, Amphoras and the Ancient Wine Trade (Agora Picture Book vi).

¹³¹ Cf. Hesperia xxxi (1962) 38 n. 7.

¹³² See Grace, , Hesperia iii (1934) 200.Google Scholar

¹³³ We rely on the published material from the excavation, Hesperia xxxi (1962), 26–61; there are frequent references to unpublished material, but descriptions are not specific.

¹³⁴ See ibid. nos. 12, 13, 15, 24, 25, 26, 35, 37, 38, 41, 43.

¹³⁵ See Hesperia xxiii (1954) 72–5; xxi (1952) 121 f.

¹³⁶ Hesperia xxiii (1954) pl. 24 (Agora P 12406); xxi (1952) 122.

¹³⁷ Hesperia xxiii (1954) 73 no. i, pl. 24a.

¹³⁸ Hesperia xxiii (1954) 73 no. 5, pl. 24e.

¹³⁹ Ibid. 73 f. Parallels may be seen also among the selection of sherds recently published from the destruction deposits in the house of Mikion and Menon in the Athenian Agora; the house is thought to have been de stroyed by fire at the end of the fourth century B.C.; Hesperia xxxviii (1969) 383ff., especially 390–4, pis. 102, 103.

¹⁴⁰ BSA lvii (1962) 75–114, especially loiff. with references to other sites on 102–3. C. Phoenix xx (1966) 3–31, J. W. Graham, ‘Origins and Interrelations of the Greek House and the Roman House’ with numerous plans.

¹⁴¹ Items of ‘farm equipment’ have been found at several ancient sites, and those not always isolated country houses; their discovery in town-sites shows that townsfolk were actively employed in agriculture and the harvesting of field crops, and underlines the close connection then between city life and an agricultural economy. For example, the Hellenistic House at the hill-town of Praisos in Crete contained oil-presses and a ‘separator’ pot for the oil (BSA viii (1901–2) 259–70, especially 264–9). Olynthian houses produced crushing wheels for pulping olives, oil-presses, and grape-treading floors (Olynthus viii. 337–42 pls. 81–3). At Merenda in central Attica trial excavations produced a press for olives or grapes associated with a house site (AAA i (1968) 31; Ergon 1960, 30–1; 1961, 37 ff.). Presses at some town houses may indicate industrial activity such as dyeing; Hesperia xxxviii (1969) 322–4 pl. 81, for a room with a press at Halieis in the Argolid. Isolated houses on regularly divided farm plots near the Greek colony of Chersonesos in the Crimea include rooms with presses, vats, etc. (our FIG. 16q for one example); these domestic finds are supplemented by remarkable remains of enclosed fields, vineyards and orchards with division walls, soil-retaining walls, and planting pits for vines and trees; see Eirene viii (Prague, 1970) 123–74, M. Dufkova and J. Pečirka, ‘Excavations of Farms and Farmhouses in the Chora of Chersonesos in the Crimea’, especially pls. 3, 5, 7, 8, 10; Ricerche storiche ed economiche in memoria di Corrado Barbagallo, i (Naples, 1970) 459–77, J. Pečirka, ‘Country Estates of the Polis of Chersonesos in the Crimea’.

¹⁴² Such have been rarely identified or closely examined. But an inscription relating to the temple estates of Delos, Rheneia, and Mykonos indicates the range of farmbuildings used on those islands and leased to tenants (IG xi 2.287A. 142–74). They included enclosures and roofed buildings for the use of the tenant and workers, for animals and for storage, e.g. οἰκία, οἴκημα, κλείσιον (with elements specified as ἀνδρώνιον, θάλαμον, ὑπερῴδιον), ἱπνών, μυλών, πιθών, ἀχυρών, πύργιον, βούστασις, προβατών; see Hesperia xvii (1948) 243–338, J. H. Kent, ‘The Temple Estates of Delos, Rheneia and Mykonos’, especially 289– 301. Two sites in southern Attica claimed as farmhouses or centres of small estates have circular threshing-floors set close by; Hesperia xxv (1956) 122–4, 124–6. The rural site at Delphinion, Chios, included a main building—the probable farmhouse—and minor structures thought to be ‘outbuildings’; BSA li (1956) 49–51. Also on Chios a farmhouse site at Pindakas, Emborio, preserved terraced foundations of Classical date and evidence of later (Roman) occupation in the form of an olive crusher and boulder weights for a press; BSA liii–liv (1958–9) 295–309, especially 296–8, 303, pl. 72. Cf. AJA lx (1956) 51–4 for farm enclosures and towers, with presses, etc., on Siphnos.

¹⁴³ Xenophon, , Memorabilia iii. 8. 8–9Google Scholar and Oeconomicus ix. 4 (quoted in BSA lvii (1952) 105 n. 40); Aristotle, , Oec. i. 6. 7–8 (1345a)Google Scholar; Aeschylus PV 450–3; Eupolis 378; all conveniently collected, quoted, and translated in Olynthus xii. 400, 410, 414, 440ff. (Testimonia Selecta nos. 2, 4.6, 60, 142, 144).

¹⁴⁴ BSA lvii (1962) 103–5; Olynthus viii. 144–6. The Priest's House at Zoster (Vouliagmeni), Attica has the same orientation with court to the south and principal rooms to the north (FIG. 16, h–j); AE 1938, 1–31, Ph. D. Stavropoulos, Ἱερατικὴ Οἰκία ἐν Ζωστῆρι τῆς Ἀττικῆς, especially 4. (fig. 4), 12 (fig. 12). At Seuthopolis in Thrace, owing to the street layout, the houses did not face the south, but sought the same basic advantage (FIG. 16c); Antiquity xxxv (1961) 91–102, especially 97: ‘The main living rooms were placed on the north-west side of the court with their backs to the north winds which blow down from the Balkan range; they were wide open on the south-east, facing the court.’ Cf. Thompson, H. A. and Wycherley, R. E., Agora xiv: The Agora of Athens (1972) 176 fig. 41Google Scholar, 181–2 fig. 44, but also for divergencies 178–9 fig. 42.

¹⁴⁵ Wiegand, Th. and Schrader, H., Priene (1904) 285–300 figs. 298Google Scholar, 316 pl. 21. Cf. AJA lxxvi (1972) 295–301, J. W. Graham, ‘Notes on Houses and Housing Districts at Abdera and Himera’; he identifies a similar house-type at Abdera in Thrace, arguing that a row of houses part excavated there should be interpreted as the street frontages of four houses (ABCD), the middle two (B, C) being complete. So interpreted, the length north—south is greater than the width, being in both cases c. 20·5 m.—‘approximately 70 Ionic feet (70×29·5 cm. = 20·65 m.)’; the width may vary, being difficult to estimate owing to incomplete excavation, but may have been c. 13 m. Graham adduces also a house at Olbia and the less regular houses at Colophon (FIG. 16k; Hspeeria xiii (1944) 91–171 pls. 9, 10), and suggests that this type was characteristic of Ionian cities of Asia Minor and their colonies.

¹⁴⁶ BSA lvii (1962) 104 fig. 12, illustrates a few more examples of the type than FIG. 16; the type included Olynthian examples (next note), and individual houses at Pella, (PAE 1914, 133–48; 1915, 237–44)Google Scholar, Eretria, (ADelt i (1915) 118, 124–31)Google Scholar, Seuthopolis, (Antiquity xxxv. 96 fig. 2 pl. 10 b, c)Google Scholar, and Peiraeus (Milchoefer, , Karten von Attika, Text i. 55–6Google Scholar).

¹⁴⁷ Olynthus viii. 55 pls. 84–5 (Villa of Good Fortune); 63 ff. pi. 87 (House of the Comedian); pls. 88–90 (Houses A-1 and A1 to A10, larger than the normal square plot with extensions running up to the city wall); pls. 95–6 (House Av6); pl. 105 (House ESH1); Olynthus xii. 259–60 pls. 222–5 (South Villa

¹⁴⁸ BSA lvii (1962) 105 n. 49 gives individual dimensions of oblong Olynthian houses.

¹⁴⁹ Hesperia xxviii (1959) 98–103; Agora xiv. 178 fig. 42. For rather larger houses in the Industrial Quarter, ibid. 176 fig. 41, and 181–2 for the still larger House of the Greek Mosaic (c. 23 × 17 m.).

¹⁵⁰ Olynthus viii pls. 95, 99, for plans of two insulae of houses with ground measurements marked, given as to centres of party walls. Pl. 95 shows Block Av: the five houses in the north row have a depth north-south of 17·15–17·20 m. and frontages of 16·93, 17·17, 17·25, 17·43, and 17·45 m.; the five southern houses have a depth of 16·85–17·15 m. and frontages of 17·24, 17·27, and 17·40, and two anomalous frontages of 25·05 m. (oblong house Av6) and 9·25 m. Pl. 99 shows block Avii, the five northerly houses with a depth of 17·08–17·15 m. and frontages of 17·10, 17·20, 17·25, 17·25, and 17·43, and the five southern houses with a depth of 17·00–17·06 m. and frontages of 17·12, 17·20, 17·25, 17·32, and 17·37 m. The house walls are generally 0·45 m. thick so that a measurement over both side walls would bring several up to a width of 17·60–17·70 m.

¹⁵¹ FIG. 16m. Wiegand and Schrader, Priene pl. 21 for several blocks of houses and fig. 298 for the much-illustrated House 33, with fig. 316 for its late peristyle form; these are reproduced, for example, in R. Martin, Urbanisme dans la Grèce antique (1956) as figs. 45, 46, 47; and House 33 on its own is figured in several general works on architecture. J. W. Graham refers to it as the ‘Musterbeispiel’ for Priene, (Phoenix xx (1966) 13)Google Scholar but notes that in having columns it is exceptional rather than typical (AJA lxxvi (1972) 296).

¹⁵² A country house at Merenda in Attica is described as c. 17·50 × 20 m. over all (AAA i (1968) 31). The House of the Greek Mosaic, just south-west of the Athenian Agora, is described as c. 17 × 23 m. (Agora xiv (1972) 181–2); of irregular shape it has a rectangular northern half, 15·0 m. wide (formed by two large rooms, one a dining-room, and a passage in between), but this width is extended to 16·0 m. by an outer east wall and expands to c. 17·0 m. across the court (the 15·0 m. width is closer to the size of the Priest's House at Zoster, 15·20 × 19·40 m.). The new Bouleuterion at Athens measured 17·50 × 22·50 m. over its foundations, but was reduced to 16·90 × 21·50 m. over its walls.

¹⁵³ House C —x 7 in section O. Olynthus viii. 28–9 pl. 5.2 suggests that the excavated part of the block represents three houses (the central element being a complete rather narrow house of 18·00 × 13·50 m. as described above, with parts of other houses on either side). Olynthus xii. 264–71 pls. 229–31, especially p. 265, reinterprets the remains as two houses, so that house C—x 7 should include rooms e, i, m, and v on its west side so as to correspond to the ‘usual Olynthian house unit of c. 17 m.’.

This argument is not convincing because (a) the total width of the house would be over 18 m.; (b) rooms e, i, m, and v have no doorway set westwards to the house, while their floors are set lower; (c) if a party wall were restored on the east side of these rooms, they would be completely enclosed, with no entry from any side. The house has an irregular internal plan, different from the pastas norm and may be anomalous. In passing, it is worth noting that its proportions (13–50 × 18.00) are 3:4.

¹⁵⁴ Olynthus viii. 29–36, 47, especially 33–4. At least one house, B i 5, was actually built with a depth of the full 60 feet rather than a depth reduced by conceding space to a drainage alley between the two rows of houses in a block. The oblong houses built against the city wall (row A) measured 70 feet from street to city wall.

¹⁵⁵ Wiegand and Schrader, Priene 50; Schede, M., Die Ruinen von Priene ² (1964) 11Google Scholar; R. Martin, op. cit. 114: ‘Avec une variation de quelques centimètres seulement, les îlots mesurent 47 m. 20 de long et 35 m. 40 de largeur, soit 160 × 120 pieds de o m. 295, unité qui se trouve identique dans le temple d'Athéna; la largeur de la rue, comme à Olynthe, n'est pas comprise dans les dimensions de I'insula.’ Cf. op. cit. 234 fig. 45 for a group of insulae, one quartered to give four uniform-size houses.

¹⁵⁶ Cf. houses at Abdera in Thrace and Himera in Sicily, n. 145 above. Graham, , AJA lxxvi (1972) 295–301CrossRefGoogle Scholar, especially 296 and 300, proposing that the Abdera ruins represent two uniform street frontages of 70 (Ionic) feet, and that the Himera houses represent a building-block laid out as a square of 100 (Doric) feet each way, quartered to give four houses each, ideally a square of 50 × 50 (Doric) feet, that is, 16–30 × 16–30 on the ‘Doric’ foot of 0–326 m., says ‘It is doubtful whether such an elegantly decimal scheme can be found anywhere ebe in ancient town planning.’

¹⁵⁷ DA (1904) s.v. Mensura; OCD² s.v. Measures.

¹⁵⁸ Ancient building specifications indicate the commoner builders’ terms of reference: IG i². 372 for the Erechtheum at Athens (c. 409–406 B.c.), specifying various sizes of quarried stone and using feet, half-feet, and palms; IG ii². 1666 for the Telesterion at Eleusis (c. 356–352 B.c.), also describing large cut stones of various type in feet, halffeet, and palms; IG ii². 1668 for the Arsenal of Philo in the Peiraeus (c. 340–330 B.c.) specifying main dimensions in plethra (100 feet), feet, and half-feet; IG ii². 463 with AJA liv (1950) 337 ff., for repairs to the city-walls of Athens (last decade of fourth century B.C.), including work in stone, mud-brick, and timber, such as rubble pillars 3 half-feet high, 5 half-feet broad, and 10 feet apart, with the brick-, timber-, reed-, and clay-work of the parapet and battlements specified in feet, brick-course heights, palms, and finger-breadths (daktyls).

¹⁵⁹ See n. 157 above. Cf. Antiquity xl (1966) 230–2, ‘Shorter units of length’, giving a foot of 0·316 m. as the ‘usual Greek foot’, an Olympic foot of 0·309 m., and an Attic foot of 0·295 m., equivalent to the Roman. Authors often disagree, e.g. Robertson, D. S., Greek and Roman Architecture (1945) 82, 149, 183Google Scholar, recognizes an Attic foot of 0·328 m. ‘used in all Athenian buildings till Roman times’, a later ‘Attic foot’ of 0·295/0·296 m. used in Roman times but found earlier in the Athena Polias temple at Priene, and presumes that the longer foot was used in Philo's Arsenal of the mid fourth century B.c. Dinsmoor, W. B., The Architecture of Greece (1950) 161, 194–5, 199, 241Google Scholar, rejects a supposed Attic foot of 0·308 m., thought to have been used in the Parthenon (‘but which no Greek ever employed’), identifies a Doric foot of 0·32685 m. for the Parthenon, of 0·326 m. for the Erechtheum, and 0·327 m. for the Propylaea, but cannot accept a foot as large as 0·328 m. in Athens, and abo recognizes a shorter foot of 0·295 m. at Priene. Plommer, H., History of Architectural Development i, Ancient and Classical Architecture (1956) 173Google Scholar considers that Dinsmoor's attempt to formulate two standards, ‘Ionic’ and ‘Doric’, involves too many discrepancies, and avoids the obvious dimensions of 100 × 125 feet for the Parthenon, i.e. an Athenian foot of 0·308 m. Boardman in The Art and Architecture of Greece 12 considers a search for exact standards unfruitful owing to the wide variety of units in use (not just two main standards), and notes that such searches prove only that a basic standard was used in a single building, with no necessary correspondence to any other, since so much depended on the mason's measure, marked off perhaps but checked with no national standard. (The latter point could certainly account for the minor discrepancies as between the three Periclean buildings on the Acropolis, but on any large-scale contract, involving several distant quarries and different gangs of men, one can presume that something like a checked standard would have to be used. Certainly the use of both a long ‘Doric’ and a short ‘Ionic’ foot would be excluded.) Carpenter, R., The Architects of the Parthenon (1970) 53, 67, 101, 114, 117, 125, 175Google Scholar, recognizes in the Parthenon elements exemplifying an older foot of 0·2957 m. and a newer Periclean foot of 0·326 m.

¹⁶⁰ Dinsmoor ibid. 54 n. 4, 222, 229; Olynthus viii. 45–51; Graham, , AJA lxxvi (1972) 295 ff.CrossRefGoogle Scholar, in studying houses at Abdera in Thrace and Himera in Sicily, accepts Dinsmoor's view that the Ionic foot was normal in Asia Minor and southern Italy (and from there was adopted as the Roman foot), but that the Doric foot was regular in Greece and the Greek colonies of Sicily).

¹⁶¹ IG ii². 1668 for the specifications; cf. Dinsmoor, ibid. 241–2.

¹⁶² It is often stated or implied that in the fourth century B.c. it was the larger, Doric, foot of 0·326–0·328 m.; cf. Robertson, ibid. 82, 183 n. 2; Dinsmoor, ibid. 241 n.; AJA liv (1950) 341, 342, 350. Hesperia xx (1951) 202 n., 207, 209, assumes that the prevalent standard before the mid fifth century B.C. was the shorter foot, when describing socles 0·45 m. thick as measuring 1½ Attic-Euboic feet.

¹⁶³ Cf. the Metrological Sculpture in the Ashmolean Museum, Oxford. It is a small pediment representing a man's head and shoulders, with arms outstretched as if to measure an orgyia (fathom) of 6 feet, and showing near the head an outline of a footprint, equivalent to the shorter foot of 0·295–0·296 m., but the arms extend to seven, not six, such foot lengths. Boardman, ibid., for remarks on its possible connection with standardization of lengths.

¹⁶⁴ Xenophon, , Memorabilia, iii. 1. 7Google Scholar; Cyropaedia vi. 3. 25.

¹⁶⁵ BSA lvii (1962) 77, 106 with references.

¹⁶⁶ Olynthus viii. 223–9; xii Pls. 49, 76, 77, 79; Antiquity xxxv (1961) 97–8.

¹⁶⁷ For well-preserved examples of an extreme kind cf. the stone-built, corbel-roofed houses in the mountains of Euboea; AJA xxix (1925), F. P. Johnson, ‘The Dragon Houses of Southern Euboea’. A hilltop house at Dystus in Euboea also displays stone walls preserved in places to upper-room levels: AM xxiv (1899) 465–6 pl. 6; Greece and Romi ² xviii (1971) 145–6. Wiegand and Schrader, Priene 302, for houses with stone walls at least up to the ceiling level of ground-floor rooms. Bruneau, P. and Ducat, J., Guide de Délos (1965) 35–6.Google Scholar

¹⁶⁸ Cf. House A in the Industrial Quarter of the Athenian Agora, with fifth- and fourth-century-B.C. walls, some in a chequerboard pattern of large blocks and small rubble (Hesperia xx (1951) 191–3 pls. 65 a, c, 66e); House B with an external wall of alternating polygonal blocks and stackwork, and again a socle of limestone blocks with stackwork (ibid. 196–7 fig. 10 pl. 67a; 209 pl. 68d); House K with stylistic contrasts in two parts of a socle, resulting not from structural need, but probably from the individual styles of two masons (ibid. 239–41 fig. 17 pl. 75). Also Athenian fifth-century houses with good outer walls of large limestone blocks, showing ‘careful polygonal jointing tending to the rectangular’ and surfaces ‘striated with short vertical strokes’, but with inner walls ‘normally of rubble masonry’ built of small blocks carefully laid in clay, and fourth-century house foundations of large rough boulders and stackwork (Hesperia xxviii (1959) 98–103 pl. 20). The House of Mikion in the Athenian Agora displays ‘a fine stretch of polygonal masonry along the street’ having a ‘nearly constant height of 0·55 m.’ as a socle for mud-brick, but topped later by ashlar conglomerate blocks, and within, ‘much cruder polygonal masonry for the walls of the first period’, but this also ‘0·50 m. high of rubble masonry’, and secondary internal partitions of light rubble (Hesperia xxxviii (1969) 385–7 pls. 101, 102). Ashlar and rubble contrasts were noted also in the walls of the Sanctuary of Demeter and Kore on the Acrocorinth (ibid. 301, 305–6 pls. 76·8). For contrast at Olynthus, , Olynthus xii pls. 12, 243.Google Scholar

¹⁶⁹ e.g. the house of Mikion and Menon at Athens (n. 168) and House Av6 and the Villa of the Bronzes, at Olynthus, (Olynthus viii pl. 30; xii pls. 202, 207, 221).Google Scholar

¹⁷⁰ p. 361 above.

¹⁷¹ Hesperia xx (1951) 193–4 for an Athenian house with a heightened socle to raise the bottom of its mud-brick superstructure to 1·60 m. above the floor of the great drain outside; ibid. 197 for an adobe wall lying ‘only 0·30 m. above the internal floor, but 1·20 m. above the street level outside’, and a wall described as ‘well above the danger of being undermined by flood’. For Olynthian examples cf. the House of Many Colours and the House of the Tiled Prothyron, with particular contrasts between inner and outer socles in houses terraced into the slope (Olynthus xii. 184, 215, pls. 158–66, 176–81).

¹⁷² Vitruvius ii. 8. 17; Olynthus viii. 214, 227–8.

¹⁷³ BSA lvii (1962) 77, 106 with refs. Hesperia xxxviii (1969) 301; Olynthus viii. 227 for the standard, with some few thinner walls, e.g. a mere partition of 0·20 m. thickness and a wall measurable as 0·31 m. between stucco faces, and some few thicker ones as in Houses A iv 1 and 3 which have an outer wall of c. 0·60 m. thickness (Olynthus xii pls. 50, 52).

¹⁷⁴ AE 1938, 4 fig. 4, 5, 7. Personal observation confirms that the main south wall has a broad foundation as in the published plan; but the west wall also was c 0·60 m. thick as shown by one or two upstanding boulders in its course extending the full width of the wall, while some of the later inner partitions were narrower at 0·40 m. and some were as narrow as 0·35 m.

¹⁷⁵ Mussche, H. F., Thorikos v (1968) 105–15 pl. 5Google Scholar, insula 10, building EA-EB-EC-ED.

¹⁷⁶ Observed at several villages in 1969–72.

¹⁷⁷ Vitruvius ii. 3. 3.

¹⁷⁸ AE 1932, 54.

¹⁷⁹ Olynthus viii. 225; xii. 225.

¹⁸⁰ ii. 3. 3–4; cf. IG ii². 1672, 56, specifying bricks 1½ feet long for Eleusis (329 B.C.).

¹⁸¹ See refs. in n. 179.

¹⁸² Cf. AJA liv (1950) 348 on horizontal timbers to be used in the parapets of the city walls of Athens according to specifications in IG ii². 463; Antiquity xxxv (1961) 96 for wooden framing in the city walls of Seuthopolis, indicated by preserved nails.

¹⁸³ Olynthus viii. 226–7, 291; Olynthus xii. 225, but ‘an interesting and rare feature is the covering of the exterior north wall with a coat of cement and lime stucco’.

¹⁸⁴ n. 183 and Olynthus xii. 215 (clay plaster with red tinting on adobe) and 219 (‘a facing of clay plaster on the stereo’); for Athens, Hesperia Supp. iv. 9, 19; Hesperia vi (1937) 18; xxiii (1954) 58; for Corinth, , Hesperia xxxviii (1969) 300–1Google Scholar, for an ‘Original clay plaster surface’, and for a ‘fine lime cement’ on the walls of the Sanctuary of Demeter and Kore on the Acrocorinth, also ibid. 308; cf. IG xi. 2. 199A, 108 (275 B.c.) on payment to a contractor for undertaking to stucco columns and to coat a peristyle with earth (mud or clay plaster).

¹⁸⁵ IG ii². 1672, 83–4, 203; Dem. iii. 29; xxiii. 208; Thuc. iii. 20; IG ii–iii². 2499, 7–8 (306/5 B.c.), an Athenian tenancy agreement specifying that Diognetus shall coat (with mud plaster or a wash) those walls requiring this treatment on property he is to occupy.

¹⁸⁶ Cf. Olynthus viii. 230–8; xii. 185–6; Antiquity xxxv (1961) 98; Hesperia xxxviii (1969) 391 for burnt timbers in a well at Athens.

¹⁸⁷ Cf. Olynthus viii. 232–8; BSA lvii (1962) 83–4,108,113.

¹⁸⁸ Cf. Olynthus viii. 249–66 pls. 69–71; xii (Testimonia Selecta) nos. 27, 28, 31, 45, 47, 56, 92, 95, 108, 144, and pp. 455, 460 for Auleia Thyra, Thyra, and Thyris.

¹⁸⁹ The Testimonia in n. 188 specify frequently the courtyard door barring the house against the outside world, also sometimes a back-door or side-door, doors to separate parts of the house, doors closing off the women's quarters, and a door for a tower storeroom, also serving as the maids' quarters.

¹⁹⁰ See p. 367 above, PLATES 68, 69a.

¹⁹¹ BSA lvii (1962) 107–8 with refs., pl. 24c; note that of the several Athenian column-bases there cited, one base, 0·36 m. in diameter, in building F in the Archaic Prytaneion complex displays a deep central mortise (Hesperia Supp. iv (1940) 23, fig. 16), and another a high drumbase, 0·30 m. diameter, displays a smaller central mortise (Hesperia xx (1951) 222 pl. 71a). Cf. Olynthus viii. 239–40 for the much commoner pillars and rectangular bases of Olynthus; again, mortíses are rare and then sometimes shallow; ‘the Olynthian builder depended on the weight of the pillars to hold them firmly in place’. In the case of modern village-houses (p. 426 above) where wooden pillars exist, they often stand on earth or rubble, sometimes on flat slabs, less often on rather higher blocks (which may be crudely shaped), but all are stabilized by sheer weight rather than secured in any form of socket.

¹⁹² BSA lvii (1962) 107 n. 71. AE 1938, 9 for the five preserved column-bases in the Priest's House at Zoster, all displaying plain flat-topped drums, 0·33–0·35 m. diameter and 0·15–0·18 high, rising from squarish blocks of 0·43 × 0·49 m. size.

¹⁹³ Olynthian pillar-bases did not generally match the thickness of partition-walls. Certainly some pillars were very sturdy (e.g. two well-preserved stone pillars, 0·51 × 0·21 m. in cross-section and 1·82 m. high) and some fully 0·425 m. square. Moreover, the existence of some thickset columns is proven by inscribed circles, 0–35 m. diameter, on two square column-bases and by surviving stone capitals, 0·31 and 0·35 m. diameter. But the average pillar was slimmer (0·40 × 0·25 m.) and some were as small as 0·22 × 0·18 m. and even 0·20 × 0·15 m. (to judge from the stone bases and capitals used with the wooden shafts). Yet these must have been adequately strong. Stone capitals also showed that some columns were as slim as 0·21 and 0·20 m. diameter. Olynthus viii. 71, 74, 119, 166, 239–48; xii. 43–4, 244–5, Pls. 34–8. 209–12.

¹⁹⁴ Timber was a costly and valued commodity—one of the most expensive items involved in house-building—and so owners and tenants were eager to remove it, and neigh bours to filch it (see n. 15 above). Good supplies of timber were probably obtainable in the hinterland of Olynthus, in Macedonia and Thrace, and could well have been used on a more lavish scale in private houses than would be the case in Attica (cf. Plato, Critias 111 b-d for a description of erosion and probable deforestation in Attica in his day, which could mean that trees large enough to provide good sturdy timbers would be in short supply; also BSA lvii (1962) 114). Even so, a gift of timber from Macedon for roofing his own house was a sufficient bribe to buy the Olynthian traitor Lasthenes; Dem. xix. 265 (343 B.C.). Furthermore, at Dystus, in the well-known house which preserves some of the arrangements for an upper floor (see n. 167 above), examination of the interior showed a contrast between the thick walls of massive blocks and the two rows of close-set small squarish beam-holes cut into them to hold the floor joists of the upper room; the builders could construct as strongly and ruggedly as they liked in stone but had to be content with rather puny timbers of small cross-section.

¹⁹⁵ Olynthus viii. 240.

¹⁹⁶ See n. 193 above.

¹⁹⁷ Cf. Orlandos, A. K., AE 1916, 94 ff.Google Scholar, figs. 15, 16, 21; Τὰ Ὑλικὰ Δομῆς (1955) i. 9 fig. 3, 10 fig. 4, 11 fig. 5, for two illustrations from vases showing slim, presumably wooden, columns with very wide, projecting flat bases and one of pole-thin columns with Doric capitals and no bases; Furtwängler, and Reichholz, , Griechische Vasenmalerei iii. 287 pl. 57Google Scholar for a pyxis (BM Cat. iii no. E 773 from Athens) with a domestic scene of women in a house courtyard, showing a half-open door and a slim column with broad spreading Doric capital and equally broad, projecting flat base; Schefold, K., Kertscher Vasen (1930) pl. 12 a, b.Google Scholar

¹⁹⁸ Olynthus viii. 247 pl. 67.1.

¹⁹⁹ e.g. a house at Drouva above Olympia with a veranda 10 m. long and 2 m. deep, formed by six posts no more than 0·10–0·15 m. square at 1·80 m. intervals, carrying a planked and railed gallery, with six upper pillars to support a heavy tiled roof.

²⁰⁰ BSA lvii (1962) 107; Olynthus viii. 281–90 pls. 15–17, 35, 42, 45, 51, 53, 84 fr.; for the late-fourth-century House of the Greek Mosaic at Athens see Agora xiv (1972) 180–2 pl. 89.

²⁰¹ Cf. Agora xiv. 175, 176, 178; Olynthus viii. 158 pls. 85–90, 97–99, 102; xii pls. 6, 14, 18, 24, 42, 106, 110–11, 175–6, 190. Olynthus has preserved nothing so massive as the Vari House pavement, though it has courts surfaced with mosaic or plaster.

²⁰² See p. 423 above.

²⁰³ See ibid, with refs.

²⁰⁴ BSA lvii (1962) 76, 80, 108–9.

²⁰⁵ Olynthus viii. 141—51.

²⁰⁶ Ibid. pls. 89, 94, 95, 97, 99 for rows of square houses, mostly with two, three, or (rarely) four rooms along their north side; one or other of these may have internal subdivisions (e.g. the kitchen complex with bathing- and cooking-cubicles at one end).

²⁰⁷ e.g. House Av 6, a very large house created by taking over part of the adjoining house-plot; ibid. pls. 95–6, and illustrated in our FIG. 16f.

²⁰⁸ Olynthus viii. 167 ff., ch. iv ‘The Rooms’; 281–2, 284–5, 291.

²⁰⁹ Ibid. pls. 87, 89 (Houses A 3, A 4, A 6), 97 (Houses Avi 5 and 7).

²¹⁰ Ibid. pl. 88, House A 10.

²¹¹ PAE 1914, 133–48; 1915, 237–44; Olynthus viii. 148–9.

²¹² FIG. 16 h, i,j, distinguish the three structural phases as interpreted in AE 1938, 4 ff., 14 f., 15 ff.

²¹³ For that reason, perhaps, upper rooms were reserved in some Athenian households for the sole or main use of the women, for privacy and security. Cf. Lysias i. 9: οἰκίδιον ἔστι μοι διπλοῦν ἴσα ἔχον τὰ ἄνω τοῖς κάτω κατὰ τὴν γυναικωνῖτιν καὶ κατὰ τὴν ἀνδρωνίτιν. Also [Dem.] xlvii. 56 (quoted in n. 232) for servant girls living in an upper room, where they stayed safe after barring the door against intruders.

²¹⁴ Gf. reconstruction views of galleried pastas houses in Olynthus viii. 99 fig. 4 pl. 73; Antiquity xxxv (1961) pl. 10c (pastas-house at Seuthopolis); BSA lvii (1962) 112 fig. 13B pl. 31 (Dema House).

²¹⁵ Olynthus viii. 216.

²¹⁶ AE 1938, 12 fig. 12, 15–17 fig. 15.

²¹⁷ Olynthus viii. 171–85.

²¹⁸ BSA lvii (1962) 76, 78, 110.

²¹⁹ BCH lxxxi (1957) 515 fig. 12.

²²⁰ Olynthus viii. 186–8 pls. 52, 97, Houses Avi 6 and 10; xii pls. 158, 160, 163.

²²¹ Antiquity xxxv (1961) 98 pl. 11a.

²²² Cf. Hesperia xxxviii (1969) 298 ff. for the sanctuary of Demeter and Kore on the Acrocorinth, preserving several benches and couches set against the walls, built variously of clay and rubble, mud and brick, and ashlar blocks with fill behind. The benches are narrower, 0·45 m. and 0·60–0·65 m. wide, while the couches are more commodious, 0·75, 0·80, and 0·90 m. wide.

²²³ Hesperia xxxviii (1969) 306–7 for a room with individual couch-lengths varying from 1·45 to 1·95 m., and a second room with two uniform couch-lengths of 1·65 m. Clearly the 1·70 m. lengths available at the Vari House would be adequate.

²²⁴ See n. 219 above.

²²⁵ BSA lvii (1962) 102 with refs.

²²⁶ AE 1938, 12.

²²⁷ See n. 222 above.

²²⁸ AE 1938, 4, 6–7; the outer, east wall of the tower is only minimally preserved.

²²⁹ See n. 141 above for refs.

²³⁰ Hesperia xiii (1944) 91 ff. Present-day village-houses in various parts of Greece show parallels to this arrangement of two doors, one over the other, with external staircases of wood or stone (or a wooden flight set on stone steps) leading to the upper door. Cf. FIG. 16k and 1.

²³¹ AJA lx (1956) 51–5; Hesperia xxv (1956) 122–46 with refs. to other sites listed, and a selection of other tower-and-court complexes, such as those on Thasos, in FIG. 7.

²³² (Dem.) xlvii. 53–6: ἀλλ᾿ ἐπεισελθόντες εἰς εἰς τὸ χωρίον (γεωργῶ δὲ πρὸς τῷ ἱπποδρόμῳ, καὶ οἰκῶ ἐνταῦθ᾿ ἐκ μειρακίου) πρῶτον μὲν ἐπὶ τοὺς οἰκέτας ᾗξαν, ὡς δὲ οὗτοι διαφεύγουσιν αὐτοὺς καὶ ἄλλος ἄλλῃ ἀπεχώρησαν, ἐλθόντες πρὸς τὴν οἰκίαν καὶ ἐκβαλόντες τὴν θύραν τὴν εἰς τὸν κῆπον φέρουσαν Εὔεργός τε οὑτοσὶ ὁ ἀδελφός ὁ Θεοφήμου καὶ Μνησίβουλος ὁ κηδεστὴς αὐτοῦ…εἰσελθόντες ἐπὶ τὴν γυναῖκά μου καὶ τὰ παιδία ἐξεφορήσαντο ὅσα ἔτι ὑπόλοιπά μοι ἧν σκεύη ἐν τῇ οἰκίᾳ.…πρὸς δὲ τούτοις, ὦ ἄνδρες δικασταί, ἔτυχεν ἡ γυνή μου μετὰ τῶν παιδίων ἀριστῶσα ἐν τῇ αὐλῇ, καὶ μετ᾿ αὐτῆς τιτθή τις ἐμὴ γενομένη πρεσβυτέρα…ἀριστώντων δὲ ἐν τῇ αὐλῇ, ὡς ἐπεισπηδῶσιν οὗτοι καὶ καταλαμβάνουσιν αὐτὰς καὶ ἥρπαӡον τὰ σκεύη, αἱ μὲν ἄλλαι θεράπαιναι (ἐν τῷ πύργῳ γὰρ ἧσαν, οὗπερ διαιτῶνται) ὡς ἤκουσαν κραυγῆς, κλείουσι τὸν πύργον, καὶ ἐνταῦθα μὲν οὐκ εἰσῆλθον, τὰ δ᾿ ἐκ τῆς ἀλλγς οἰκίας ἐξέφερον σκεύη.…. Ibid. 63: ὁ δ᾿ Εὔεργος οὑτοσὶ εὐθὺς ἐκ τῆς πόλεως μεθ᾿ ἑτέρων ὁμοίων αὑτῷ ἐλθὼν εἰς ἀγρόν, τὰ ὑπόλοιπα σκεύη, εἴ τινα τῇ προτεραίᾳ ἐν τῷ πύργῳ ἦν καὶ οὐκ ἔτυχεν ἔξω ὄντα, ἐπειδὴ δ᾿ ἐγὼ ἦλθον, διὰ τὴν χρείαν κατηνέχθη, ἐκβαλὼν τὴν θύραν ἥνπερ καὶ τῇ προτεραίᾳ ἐξέβαλον κακῶς ἐνεστηκυῖαν, ᾤχετό μου λαβὼν τὰ σκεύη.

²³³ Lucian Timon 42: Ѡ δίκελλα καὶ φιλτάτη διφθέρα, ὑμᾶς μὲν τῷ Πανὶ τούτῳ ἀναθεῖναι καλόν αὐτὸς δὲ ἤδη πᾶσαν πριάμενος τὴν ἐσχατιάν, πυργίον οἰκοδομησάμενος ὑπὲρ τοῦ θησαυροῦ μόνῳ ἐμοὶ ἱκανὸν ἐνδιαιτᾶσθαι, τὸν αὐτὸν καὶ τάφον ἀποθανὼν ἕξειν μοι δοκῶ. Lucian sets the scene in the foot-hills of Hymettus, ibid. 7: τίς οὖτός ἐστιν, ὦ Ἑρμῆ, ὁ κεκραγὼς ἐκ τῆς Ἀττικῆς παρὰ τὸν Ὑμηττὸν ἐν τῇ ὑπωρείᾳ… Cf. Pausanìas i. 30. 3–4, placing Timon's tower north-west of the city near the Academy: ᾿Ακαδημίας δὲ οὐ πόρρω Πλάτωνος μνῆμά ἐστιν…κατὰ τοῦτο τῆς χώρας φαίνεται πύργος Τίμωνος. For towers (πύργοι and πυργίδια) in the Attic countryside in Roman times see IG ii². 2776, 15, 24, 65, 115, citing examples in the Mesogaion.

²³⁴ For two farms belonging to the Delian temple estates on Rheneia, viz. Rhamnoi and Charoneia, each with a tower included in the inventory of farm-buildings, see IG xi. 2. 287A, 154–5: Ῥάμνους … καὶ παρέλαβεν θύραν αὐλείαν, προβατῶνα ἄθυρον, πυργίον τεθυρωμένον, ἀνδρώνιον τεθυρωμένον, κλείσιον τεθυρωμένον, βούστασιν ἄθυρον, θαλάμους ἀθύρους, ἀμπέλους … συκὰς … ῥοήν. Ibid. 164—9: χαρώνειαν…παρἐλαβον κλείσιον τεθυρωμένον, θαλάμους δύο τὸν ἕνα τεθυρωμένον, βούστασιν ἄθυρον, προβατῶνα ἄθυρον, οἴκημα ὀροφὴν οὐκ ἔχον, πύργον τεθυρωμένον, θύραν αὐλείαν, ἄλλην οἰκίαν θύραν αὐλείαν, κλείσιον τεθυρωμένον, θάλαμον τεθυρωμένον, ἄλλον θάλαμον, ἀνδρώνιον τεθυρωμένον, ἰπνῶνα ἐστυλωμένον τὴν δοκόν, βούστασιν ἄθυρον, ἀχυρῶνα ἄθυρον, προβατῶνα ἄθυρον, συκᾶς … ἀμπέλους … Cf. Hesperia xvii (1948) 243 ff. J. H. Kent, ‘The Temple Estates of Delos, Rheneia, and Mykonos’, especially 251, 295. Towers are rare on Rheneia; only the two above-named farms preserved a πύργος or πυργίον Charoneia was a twin estate, having two farmyard enclosures set together, one of them preserving in one corner a house with cistern basement and in the other the collapsed ruins of a high tower, 12 m. square. Perhaps significantly neither of these farms included in its list the ὑπερῴδιον ascribed to many other farms. The special mention of such a ‘small upper room’ suggests (a) that the other buildings were ground-floor compartments (not necessarily separate buildings, but perhaps rooms set in a row or in a complex) and (b) that the ὑπερῴδιον itself was a building of limited extent which, while much lower than a real πύργος, could have been the one upstanding element in the farmyard, rising above a warren of (? stone-built, flat-roofed) huts. If so, the effect may have been rather like the ‘mini-towers in Cyprus’ illustrated in PLATE 81 a. For a farm tower used for storage in Iacinthus see IG xii. 5, 872. 52: …καὶ τοῦ πύργου καὶ τοῦ πιθῶνος τοῦ ἐν τῷ πύ[ρ]γῳ κ̣αὶ τοῦ κεράμο[υ τῆς στέγ]ης τὸ τέταρτομ μέρος … Biblical texts suggest also that towers were connected with farms and vineyards over a very long period in the Near East: Isaiah v: 1–2: ‘Now will I sing to my well-beloved a song of my beloved touching his vineyard. My well-beloved hath a vineyard in a very fruitful hill. And he fenced it, and gathered out the stones thereof and planted it with the choicest vine and built a tower in the midst of it and also he made a wine-press therein, and he looked to it that it should bring forth grapes, and it brought forth wild grapes’; Mark xii: 1: ᾿Αμπελῶνα ἐφύτευσεν ἄνθρωπος, καὶ περιέθηκε φραγμόν, καὶ ὤρυξεν ὑπολήνιον, καὶ ᾠκοδόμησε πύργον, καὶ ἐξέδοτο αὐτὸν γεωργοῖς, καὶ ἀπεδήμησε. For a smart town house of Hellenistic date (c. 125–88/69 B.C.) on Delos, built on a narrow plot but extending up into a tower of two floors, restored as decorated with pediment, triglyph friezes, and wide window arcades, see Bruneau, and Ducat, , Guide de Délos (1965) 117–19Google Scholar with drawing in fig. 21.

²³⁵ Vitruvius ii. 8. 17: ‘leges publicae non patiuntur maiores crassitudines quam sesquipedales constitui loco communi; ceteri autem parietes, ne spatia angustiora fierent, eadem crassitudine conlocantur. Latericii vero, nisi diplinthii aut triplinthii fuerint, sesquipedali crassitudine non possent plus unam sustinere contignationem.’ Pliny, , NH xxxv. 169Google Scholar comments on strength and longevity of earth walls made by packing mud between shuttering (as with modern concrete), and adds, ibid. 173: ‘Romae non fiunt talia aedificia (i.e. of mud brick), quia sesquipedalis paries non plus quam unam contignationem tolerat, cautumque est, ne communis crassior fiat, nee intergerivorum ratio patitur.’

²³⁶ Clearly seen in AM xxiv (1896) 458 ff. pl. 6.

²³⁷ Olynthus viii. 227.

²³⁸ Ibid. 227–8, citing a modern formula for calculating the required thickness of wall for any given height; by this rule, 0·35 m. thickness would suffice for house-walls meant to carry an upper floor. In modern brickwork a 10-inch or 12-inch cavity wall (c. 0·30 m.) is acceptable for such a house.

²³⁹ A house at Phloka near Olympia, measured and recorded in 1971, but in 1972 demolished and rebuilt in concrete and modern materials.

²⁴⁰ Olynthus viii. 240 f., 276–9; BSA lvii. 113.

²⁴¹ Olynthus viii. 236–8; BSA lvii. 113.

²⁴² IG i². 778, 784–8 for the inscriptions of Archedamus (Archedemus) of Thera in the Cave of Pan uphill from the house site. Cf. AJA vii (1903) 289 ff.; Jeffery, L. H., The Local Scripts of Archaic Greece (1961) 319, 323Google Scholar, noting that the alphabet is not Attic, but presumably Archedemus' local script, though not in its most archaic form, and further dating the inscriptions to c. 450–400 B.c. If that date can be accepted for Archedemus' activities at the cave site, it is interesting that there is some evidence for occupation also within that same period at the house site. Once again, fancy nearly impels us to venture beyond hard evidence, and to think of that nympholept, groping away for hours by lamplight in the dark cave, embellishing it after his fashion, carving steps and niches, planting some form of garden, and adding his signature in script and sculpture, and to wonder if he was not the same person as the anonymous first resident at our house site.

²⁴³ The excavators of the Vari Cave House would like to express gratitude to Mr. Geroulanos for writing this article at their request, for taking the photographs shown in PLATES 83–5, and for allowing part of his results to be published in the BSA so as to appear alongside their article on the Vari Cave House, and thus provide an illuminating parallel to some of their finds. They are deeply obliged to him for his kindness in showing them his collection at Trachones, for discussing the problems of beehive pots with them in 1966 and 1969—which was all the more valuable as he is both an excavator and a practical beekeeper of many years' standing—and for permitting J. E. Jones to measure the Trachones pots and prepare the drawings illustrated in FIGS. 18–20. (J. E. J.)

²⁴⁴ Trachones itself and some surviving traces of ancient farming in its area are mentioned in Bradford, J., Ancient Landscapes (1957) 33, pls. 8–10Google Scholar; pp. 29–34 deal with the remains of terraces, field and trackway systems on the coastal plain along the eastern flanks of Mt. Hymettus, with a sketch-map showing Trachones on p. 30, fig. 2. Trachones marks the site of the classical deme of Euonymon. (J.E.J.)

²⁴⁵ One possible use of the vase in an upright position, not to be absolutely excluded, might have been for the conversion of wine into vinegar, since the incisions would retain some quantity of the yeast after draining off the wine. But it is difficult to see, in that case, why only part of the interior was roughened rather than the whole, which presumably would have facilitated such an operation.

²⁴⁶ Mr. Geroulanos has informed us by letter (21.10.1972) that modern copies of the Trachones beehives have been specially made and are ready for experiments to be conducted in the spring of 1973 at Trachones and at Anavryta (near Maroussi). He notes: ‘We know now that there are two kinds of bees. One constructs the honeycombs along the length of the ceramic beehive, the other just at right angles to the length. People from Antiparos tell me that it is impossible to change the way in which bees organize their combs.—As to “smoking” the bees, I am sure that it is possible to get the extension rings off without smoking at all, unless the bees are of some very aggressive variety and there is a very strong and dry north wind blowing.’

In this connection we can cite the remarks of Mr. Thanos Michaelis of Limassol in Cyprus, who is at present (1972–3) a student at University College, Bangor, but who has had practical experience as a beekeeper with the very kind of horizontal pipe-like ceramic beehive used in Cyprus and illustrated in PLATES 79c and 80a. He informs us that the lid-ends of the pipe-hive are usually made of wood but are smeared with clay or mud for sealing and insulation; both lid-ends are removable at need, but only one, the ‘front’ lid, is provided with an elliptical flight-hole in its outer edge (illustrated in PLATE 80a.) His experience is that the bees form combs right from the rear end of the pot and always at right angles to the length of the pot; these combs are not fully circular and do not fill the whole circle of the pot (as would wads in a cartridge-case), but some hang down from the ‘roof’ of the pot while a few are even built up from its floor; and the combs are built in a series, rather like lobes, from the rear end on towards the front end of the pot (PLATE 85 c, d). The advantage of the double-ended pipe form is that the rear lid can be removed, and that with less reaction from the bees; it is from the rear end that harvesting of the honeycomb is usually done, and a few combs at a time are cut out with a long knife. Those combs not in a prime condition can be thrust back into the hive or left outside near by as food for the bees themselves. (J. E. J.)

We owe to Dr. Eva Crane of the Bee Research Association (by letter dated 8 March 1973) a furrner reference to horizontal hives still in use in Malta. The association had acquired some year sago an example of a Maltese horizontal hive, cylindrical in shape, with one closed and rounded end; it measured 17 inches in length (c. 0·43 m.) and 9¾ in. in diameter (c. 0·25 m.). It has just now acquired an example of one of the short cylinder extensions used with such hives; this is from a larger hive, being 12½ in. in diameter (c. 0·32 m.), comparable in that respect to many of our ancient Greek examples; but this modern extension or ‘eke’ is much longer than the ancient narrow rings, for it is a cylinder 15 in. long (c. 0·38 m.). Further, it has two short lugs or ‘ears’ at one end on the exterior, one on either side, for carrying the ‘eke’ in a vertical position. Local Maltese beekeepers assured Mrs. Crane's informant that the extensions are used together with the hives in a horizontal position (basically as in Mr. Geroulanos's reconstruction of the ancient Trachones example) and that the combs are built across the cylinders, ‘i.e. in the direction that would make the use of your excavated short cylinders entirely sensible’.

²⁴⁷ The site can be conveniently reached by turning off the Sounio–Laurion coast-road at c. 1−1½ km. from Sounion, where there is a bus-stop just beyond a small white church and by a large roadside restaurant (the SYRTAKI); one follows a side-road signposted as to Agia Paraskevi, and this leads uphill and links with the ancient terraced road passing the site.

²⁴⁸ Young, , Antiquity xxx (1956) 94–7 pl. 5.CrossRefGoogle Scholar The road seems in places 4·20–4·70 m. wide and the terrace-wall, constructed of large boulders, rises to a height of 3–3·60 m. in the centre of the loop, north of the Princess' Tower site.

²⁴⁹ Young, , Hesperia xxv (1956) 122–4 Pl. 34.CrossRefGoogle Scholar The site is marked on Curtius and Kaupert, Karten von Attika, sheet xv (Sounion) and mentioned by Milchhoefer in the Text iii–vi (1889) 29.

²⁵⁰ The tower must have been very much damaged sinceMilchhoefer's visits, for his description (n. 249 for ref.) mentions a height of up to 2 m. and also stubs of internal walls projecting opposite the doorway as if for radial partitions. ‘Südöstlich daneben befindet sich, begrenzt durch Mauerzüge, die vielleicht von einer viereckigen Einfriedigung herrühren, der immer noch gegen 2 Meter hohe Rest eines Rundbaues, unregelmäβig doch sorgfaltig gefügt aus mäβig groβen Steinen von 0·90 m. Mauerdicke. Im Volksmunde ist es das Denkmal der ‘Königstochter’ oder “Königin” (Vassiló poula, Vassilikula). Eine Thüröffnung von 0·88 m. Breite blickt nach Osten. Der Innendurchmesser beträgt 4 m. Der Boden ist mit Schutt und Steinen bedeckt. Dem Eingang gegenüber springt rechts und links je ein antenartiger Ansatz wie von radialen Zwischenmauern vor. Ob auch diese Anlage sepulkralen Zwecken gedient habe, ist mir sehr zweifelhaft geworden.’

²⁵¹ Young, ibid. 124: ‘The buildings then seem to be connected in some way with ancient farming in the level valley below, which is sown to grain today by farmers from Keratea’; ibid. 141: ‘The only area in Sounion today extensively planted with grain is in the lower Agrileza Valley, which lies completely outside the mining area; precisely here stands the Princess' Tower, and there can be no doubt that the chief occupation of this estate was the farming and milling of grain. The grain threshed on the circular floor outside the enclosure wall may have been milled on the ground floor of the tower, the flour stored on the upper floors.’ One might note that the concentration of hundreds of slave and free labourers in the mining industry in these hills would provide a direct incentive to efforts to produce as much food as possible in the immediate vicinity.

²⁵² Strabo ix. 1. 23 (= C 399–C 400): ὁ δ᾿ Ὑμηττὸς καὶ μέλι ἄριστον ποιεῖ.…τοῦ δὲ μέλιτος ἀρίστου τῶν πάντων ὄντος τοῦ ᾿Αττικοῦ, πολὺ βέλτιστόν φασι τὸ ἐν τοῖς ἀργυρείοις, ὃ καὶ ἀκάπνιστον καλοῦσιν ἀπο τοῦ τρόπου τῆς σκευασίασς. Cf. Pliny xi. 15. 45: ‘cum eximantur mella, apes abigi fumo utilissimum, ne irascantur aut ipsae avide vorent. fumo crebriore et ignavia earum excitatur ad opera, nam nisi incubavere, favos lividos faciunt. rursus fumo nimio inficiuntur, quando iniuriam celerrime sentiunt mella vel minimo contactu roris acescentia; et ob id inter genera servatur quod acapnum vocant.’

²⁵³ pp. 438 ff. above for the Vari Cave House; its tower had walls 0·66–0·70 m. thick, while those of the Princess' Tower were 0·82–0·90 m. thick. Young, ibid. 135, suggested a height of about 13 m. for the latter tower and cites heights of 15–24 m. for a number of better-preserved towers.